Recently, we proposed that fatigue is caused by hypoactivity within the brain’s arousal network,

refected by a fundamental slowing of processing speed in PCS patients [12]. In particular, the variance
of (visual) processing speed in a laboratory task was explained by a neurophysiological measure of
central nervous arousal, i.e., pupillary unrest, and the level of subjective mental fatigue. Thus, based
on our data, processing speed not only seems to represent a reliable measure of the brain’s activation
level, but it can also provide an objective proof of the subjective feeling of mental fatigue. Although
these results may represent an important step towards a better understanding of the neurocognitive
defcits in PCS patients, a number of questions arise from this preceding study: frst, it remains unclear
whether a reduction of processing speed can be evidenced in standard neuropsychological
assessment procedures as well. Given the high prevalence of PCS, the suitability of conventional,
clinically established tests for disclosing processing speed defcits would be of great practical
signifcance. Second, in our preceding study [12], we were not able to assess whether cognitive
domains other than processing speed are also afected by the underlying arousal defcit. A
demonstration that processing speed is selectively impaired in PCS patients would bear strong
support for our hypoarousal model. Finally, it remains unknown whether hypoarousal is a temporary
and transient, or a persisting problem in PCS patients. This is an important question for estimating the
probability whether PCS patients are able to re-uptake their job in the near future. However, while
subjective complaints of fatigue are known to prevail even after 2 years [25], follow-up observations
including an objective measure of fatigue are still lacking. Hence, longitudinal assessment of
processing speed is urgently required. Thus, the frst aim of the present study was to analyze whether
a reduction of processing speed in PCS patients is evidenced in conventional neuropsychological tests
as applied in a standard clinical setting. We compared PCS patients with cognitive complaints to
sociodemographically matched healthy control participants and used tonic alertness, measured in a
clinically established simple-response task, as our primary outcome. The second aim was to assess
whether reduced processing speed is a selective defcit in PCS patients. We employed a computerized
reaction timebased neuropsychological assessment procedure and a comprehensive test battery to
explore whether the overall cognitive profle across diferent relevant cognitive domains would be
indicative of particular defcits in tasks reliant on fast information processing speed. As secondary
outcomes, we explored whether speed-dependent measures in general, such as reaction times and
test scores relying on task completion times, were compromised in PCS patients. Additionally, we
investigated whether test scores less dependent on processing speed were less afected. Third, to
corroborate the relationship between hypoarousal and fatigue, we investigated whether performance
in the primary outcome tonic alertness simple-response measure and the other measures that
tapped processing speed was related to the degree of experienced fatigue. As a fourth aim of our
study, we included a follow-up assessment after 6 months to evaluate whether processing speed
defcits persist or ameliorate over time. To our knowledge, this is the frst longitudinal study that
particularly addresses processing speed, i.e., alertness dysfunction in PCS patients with cognitive
complaints. Methods General procedure Eighty-eight patients, fulflling the NICE criteria for PCS [20],
with subjective cognitive dysfunction after polymerase chain reaction confrmed SARS-CoV-2 infection,
who presented for the frst time from January 21, 2021 to September 27, 2022 at the Neuro-Post-
COVID-Centre of the Department of Neurology of Jena University Hospital (JUH), were prospectively
included in the study (Fig. 1). Patients underwent comprehensive clinical examination by a
neurologist in order to confrm the PCS NICE criteria and to exclude alternative medical explanations
for cognitive dysfunction before inclusion. To determine if the patient sample was adequately sized
for longitudinal analysis, a sample size calculation using G*power 3.1.9.6 [7] was conducted